Maternal, child, and socioeconomic characteristics were different for families who enrolled relative to families who did not enroll, with factors associated with enrollment differing between families who were non-responsive to recruitment attempts and those who declined enrollment.
Key Findings
Results
Overall enrollment in the Omega Tots trial was low, with 316 families enrolling compared to 1089 who declined and 1081 who were non-responsive.
Children eligible for the trial were born at <35 completed weeks' gestation and were 10-16 months of age at recruitment.
Total contacts: 316 enrolled, 1089 declined, 1081 non-responsive.
The primary outcome was whether the family enrolled, declined, or was non-responsive to recruitment efforts.
Log-binomial regression was used to calculate risk ratios (RR) for enrollment outcomes.
Results
Enrollment was negatively associated with caregivers being married, identifying as White, and children being born at later gestational ages.
Married caregivers: RR = 0.76, 95% CI: 0.62, 0.94.
White racial identification: RR = 0.76, 95% CI: 0.60, 0.94.
Later gestational age (per 1-week increase): RR = 0.96, 95% CI: 0.92, 0.99.
These associations were for enrolling rather than not enrolling overall.
Results
Enrollment was positively associated with children weighing less than 1500 g at birth, attending a neonatology specialty clinic, family participation in WIC, and living in an urban zip code.
Family participation in WIC: RR = 1.39, 95% CI: 1.13, 1.72.
Living in an urban zip code: RR = 1.68, 95% CI: 1.30, 2.17.
These associations were for enrolling rather than not enrolling overall.
Results
Factors associated with enrollment differed depending on whether non-enrolling families had declined participation or were non-responsive to recruitment.
Varied associations were identified when comparing enrolling versus declining, enrolling versus non-responsive, and declining versus non-responsive families.
The distinction between declining and non-responsive families revealed additional differences in associated characteristics.
The authors note this variation has implications for understanding barriers to trial participation.
Conclusions
The authors concluded that recruitment initiatives tailored to ensuring enrollees reflect the source population may improve generalizability of clinical trial findings.
The study hypothesized that enrolled families would not differ from those who did not participate, but this hypothesis was not supported.
Eligibility data were abstracted from medical records and linked with the child's birth certificate.
Differences in maternal, child, and socioeconomic characteristics between enrolled and non-enrolled families were identified.
The findings suggest the enrolled sample may not fully represent the broader population of preterm toddlers.
Boone K, Miglin A, Taylor P, Klebanoff M, Keim S. (2026). Factors associated with enrollment in a randomized clinical trial of docosahexaenoic supplementation in toddlers born preterm.. Clinical trials (London, England). https://doi.org/10.1177/17407745251387983